Research Summary

Output and asset price fluctuations

by Diego Comin

Description

What are the sources of business cycles? How are these shocks propagated in the economy? Why are their effects so persistent? How can we explain asset price fluctuations? How are shocks transmitted internationally?To study these questions, I have developed a series of macro models where the technology available for production is endogenous. In booms, when aggregate demand is higher, agents find more profitable to invest in improving their technologies. As a result, the business cycle affects the rate at which technologies improve the technology. Modeling this mechanism yields many interesting consequences. First, it provides a theory for medium term fluctuations in TFP. Second, since temporary macro shocks lead to persistent deviations in the level of technology relative to trend, these type of models generate lots of endogenous persistence. This feature of the model is important, for example, to explain why Japan experienced a lost decade during the 1990s despite the facts that the shocks that hit the Japanese economy at the beginning of the decade were not nearly as persistent.

Standard macro models have a difficult time in explaining asset price volatility. In those models, the stock market is given by the value of installed capital which does not fluctuate much. Once we endogenize the level of technology, the stock market is also affected by fluctuations in the value of current and future producers of the goods that embody the technology. Since profits are very pro-cyclical, and their present discounted value is very volatile, the stock market generated by my models can replicate the statistical properties of the stock market in the data. In addition, volatility and persistence of dividends are consistent with the data. In particular, the model does not rely on highly volatile counter-cyclical risk premia to explain asset prices. Rather, the macro model generates a small persistent component in dividends. We find evidence of the presence of such persistent component in the data.

The endogenous diffusion of technology is important also to explain the co-movement between developed and developing countries. In particular, since technologies diffuse from rich to poor countries, cyclical variation in the speed of technology diffusion will affect the medium term fluctuations in developing countries. This mechanism can explain, for example, why high frequency fluctuations in the US lead medium term fluctuations in Mexico.